Three new species, Eubacterium nodatum, Eubacterium timidum, and Eubacterium brachy, were described, primarily from subgingival samples taken from patients with moderate and severe adult periodontitis. Except for the isolation of E. brachy from a pleuropulmonary infection, these species have not been reported from other infected body sites. We report on the isolation of these species and an undescribed group (D-6) of asaccharolytic eubacteria also found in periodontal disease from numerous different sites of infection, mostly the head and neck. A similarity in cellular morphological properties of E. nodatum and Actinomyces sp. was noted previously. Additional similarities, particularly to Actinomyces israelii, that we found are the formation of molar tooth colonies and the isolation from cases of lumpy jaw and from the genital tract of women in association with the use of an intrauterine contraceptive device. E. timidum and E. brachy did not occur more often from any particular site outside of the head, neck, and respiratory tract. The group D-6 strains came from a variety of sites in the trunk and pelvis. These species are all obligately anaerobic, asaccharolytic, and generally nonreactive, and they grow poorly and slowly on media commonly used to isolate anaerobic bacteria. L-Lysine (0.5%) markedly stimulated the growth of E. nodatum and, to a lesser extent, another acetate- and butyrate-producing group, Eubacterium sp. group D-6, but we did not find comparable stimulants for the other species. We found the production of phenyl acetate to be a helpful marker in the identification of E. timidum and Eubacterium sp. group D-6. Although the isolation and identification of most of these species remain somewhat difficult, the evidence from dental infections and the present report suggests that these species are potential pathogens that are likely to be overlooked in infected clinical material without special attention to more prolonged incubation and use of enriched isolation media.

Certain species or subspecies of anaerobic bacteria are isolated with higher frequency from female genital tract infections than from other anatomic sites. To gain susceptibility data more specific to the treatment of these infections, nine antimicrobial agents were tested by an agar dilution technique against 230 anaerobic bacteria isolated solely from obstetric and gynecological infections. These genital isolates were, in general, very susceptible to imipenem (most active, inhibiting all gram-negative rods at s1 ,ug/ml), clindamycin (all isolates inhibited at s4 ,ug/ml), metronidazole (all gram-negative rods inhibited at s4 ,ug/ml), and chloramphenicol. Penicillin G had generally low activity against Bacteroides spp., not restricted to just the Bacteroides fragilis group, although it was very active against gram-positive species. Bacteroides bivius, a species uniquely common in female genital infections, was particularly resistant (90% MIC, 64 U/ml). Also, the Bacteroides melaninogenicus isolates were less susceptible than previously reported for isolates not

Thirteen antimicrobial agents and beta-lactamase inhibitor combinations were tested simultaneously for their in-vitro activity against a range of anaerobic Gram-negative bacilli with a standard reference agar dilution method. Overall, metronidazole, imipenem, ampicillin/sulbactam, ticarcillin/clavulanic acid and cefoperazone/sulbactam, followed by clindamycin, cefoxitin, and piperacillin, had the greatest activity. Cefotetan, ceftizoxime, and cefoperazone were moderately active, while ampicillin and penicillin were least active. Metronidazole was the only drug active against all strains, but only one strain was resistant to imipenem. Resistance was highest among certain members of the Bacteroides fragilis group, but was observed also among numerous other Bacteroides species. beta-Lactamase was produced by 94% of strains in the B. fragilis group, and by 64% of strains overall. The activities of clindamycin and cefoxitin were compared with those in previous surveys since 1982 at our institution. No clear evidence of increasing resistance was demonstrated, but the data emphasized the significant effects resulting from variations in susceptibility testing.